Method

Participants and design. The study used a 2 × 2 repeated measures design with the factors test

type (retention vs. transfer) and test format (visual representation vs. text). Participation was restricted to 18 to 30-year-old native speakers of German with no or low knowledge concerning arm anatomy. In addition, participation was limited to female students. Given the sample size, we

did not want to introduce variance potentially arising through gender differences (see, e.g., Wong, Castro-Alonso, Ayres, & Paas, 2015; for a meta-analysis, see Castro-Alonso, Wong, Adesope, Ayres, & Paas, 2019). Thirty-five female participants completed the experiment. We chose this sample size based on an estimated effect of ηp2 = .08 (power = 0.80, α error probability = .05)

using G*Power (Version 3.1.9.2; Faul, Erdfelder, Lang, & Buchner, 2007; Faul, Erdfelder, Buchner, & Lang, 2009). All participants were students of Media Communication, Computer Science and Communication Science, or Media and Instructional Psychology and participated in partial fulfillment of course requirements. The web-based experiment was created using SoSci Survey (Version 2.6.00-i; Leiner, 2018; https://www.soscisurvey.de). Participants sat in a lab room containing ten desktop PC workstations in individual cubicles.

Learning materials. The learning materials used in the learning phase consisted of an interactive

anatomy diagram on the left side of the screen accompanied by a short text next to it on the right. Labeled line drawings of the human arm (grounded in information presented in Bammes, 2009, and Tillmann, 2016) created using Inkscape (Version 0.48.4 r9939; Inkscape’s Contributors, 2012) and Versions 4.10.6033.29846 and 4.21.6589.7045 of Paint.NET (2016, 2018) were employed in the learning phase (see Figure 1). The learning targets in the visual part of the task were two bones and five muscles (or muscle groups) presented as schematic renderings of the arm bones with lines indicating the position and points of attachment of the muscles to the bones (flexors in red, extensors in blue). The images contained arm contour lines and reduced forms of shading. Participants could swap between two views of the arm by clicking on the image (see Figure 1A & 1B) using a JavaScript-based interaction.

Figure 1. Learning materials used in Experiment 1 (grounded in information presented in Bammes,

2009, and Tillmann, 2016). Panels A and B show the arm in the two poses that participants could switch between by clicking on the image.

The top of the page of the learning phase stated that participants should memorize the information contained in the image and the text. They were also instructed to change the orientation of the arm by clicking on the image. A countdown of the remaining time was displayed above the learning materials. The learning text (based on information conveyed in Bammes, 2009, and Tillmann, 2016) beside the interactive drawing contained additional information concerning each bone and muscle (group) presented in the drawing and had a word count of 97. These facts included the function and position of the anatomical components of the arm.

Learning tests. Learning performance was tested using four tests that were presented on four

consecutive pages in randomized order. The visual retention test (see Figure 2A) depicted the arm in an in-between pose between the two poses presented in the graphical display contained in the learning phase. Therefore, participants who had not clicked the image were able to complete this test. All items on the drawing used in the test were labeled using a letter. The participants’ task was to assign labels with the names of the items to letters below the drawing in accordance with the labels in that image. The seven-item visual retention test had a McDonald’s ω (McDonald, 1999) of 0.76 (throughout this article, McDonald’s ω is used for learning tests due to numerous

advantages of this reliability measure as outlined by Dunn, Baguley, & Brunsden, 2014). The text- based retention test featured seven multiple choice questions, each with seven alternative answers of which only one per question could be selected using drop-down menus. These questions asked participants for facts mentioned in the text portion displayed during the learning phase. This test resulted in a McDonald’s ω of 0.73. Transfer performance was measured using a visual task and a text-based test. The visual transfer task (see Figure 2B) showed the human arm in a pose very different from the one presented in the learning phase and thus required participants to mentally rotate the information they had previously learned. As in the visual retention test, participants were instructed to drag labeled boxes into a letter-based list in accordance with the labels shown in the illustration. The seven-item visual transfer test had a McDonald’s ω of 0.67. Finally, the text-based transfer test consisted of seven multiple-choice questions, again with only one correct answer to choose out of seven alternatives per question. This transfer test contained questions asking participants to select a muscle (group) or a bone based on written descriptions providing such information as surrounding anatomical items or specific rotations of the arm. A translated example question would be “Which muscle group is positioned at the lateral side of the lower arm and inserts at the hand bones?” These questions were designed to require participants to mentally simulate the location and orientation of the anatomical parts of the arm. McDonald’s ω of the text-based transfer test was 0.27.

Figure 2. Visual learning tests used in Experiment 1. A: Visual retention test showing the arm in a

position between the two poses used during the learning phase. B: Visual transfer test showing the arm in a pose strongly differing from the learning phase.

At the top of the four test pages, participants were informed that there was no time limit and were asked not to use any additional assistance. In terms of scoring, participants were awarded one point for each correctly selected muscle (group) or bone. Each of the four tests had a maximum score of seven.

Surveys. On each of the four learning test pages, we included items to measure the mental effort

and perceived difficulty of the respective test page (following Prisacari & Danielson, 2017). In line with Prisacari and Danielson (2017), we adapted Paas’ (1992) question item of mental effort. We based our adaptation on the German translation of this item by Syring and colleagues (2015). Prisacari and Danielson (2017) used the approach developed by Bratfisch, Borg, and Dornič (1972) to measure perceived difficulty. This approach was advanced by Marcus, Cooper, and Sweller (1996). We adapted the German translation of Marcus and colleagues’ (1996) difficulty item developed by Syring and colleagues (2015). Both question items were developed as 9-point Likert scales (see Paas, 1992 and Marcus et al., 1996, respectively). However, for both items, we used German translations of the scale endpoint labels presented by Prisacari and Danielson (2017).

Procedure. After providing informed consent, participants were asked to provide information

concerning whether they are within the age range of 18 to 30 years, are native speakers of German, have little or no knowledge of arm anatomy, and have not previously participated in the study. Furthermore, they were asked to indicate their course of study and their gender.

Next, participants were presented with the instructions for the following learning task. They were informed that their task was to memorize information concerning the human arm and that the next page contained a labeled drawing of the arm as well as a text. In addition, the page stated the time limit of 90 s. Participants received an explanation that they could change the rotation of the arm by clicking on the image. In order to prevent them from accidentally skipping this instruction page, participants had to wait 15 s before being able to proceed to the next page. Thus, participants were informed that they should click on the button labeled “Next” when they were ready, but that the button would only appear after 15 s.

The learning materials were presented on the next page, with an automatic redirect to the following page after the learning time of 90 s had passed. The following page consisted of a filler task in which participants were given 60 s to sort the 16 German federal states according to their population. After 60 s had passed, they were redirected automatically to the following page. At this stage, the four learning test pages were displayed. Participants were notified of the finalization of data collection and after their confirmation, they were thanked and provided with additional information concerning the study. Each session lasted approximately 20 min.